Although we know a great deal about language comprehension, we know much less about how it is implemented in the brain. This proposal tests two hypotheses about how verbs and their arguments (e.g. their subjects and objects) are processed in the brain, which have important implications for which aspects of verb-argument processing might be preserved vs. impaired among adults with aphasia. This research has three aims: (1) To increase our understanding of the multiple brain systems that work together to accomplish verb-argument processing, and how their situation in the brain might relate to their different functions (coarse- vs. fine-grained processing; prediction vs. integration); (2) To test our hypotheses' predictions that adults with aphasia will be less likely to predict upcoming words in general, but will still generate a special class of argument-specific predictions if they have a undamaged right hemisphere (RH) or posterior left hemisphere (LH); and (3) To understand the consequences that distributing language processing across multiple brain systems has on verb-argument comprehension. The proposal includes five experiments. Experiment 1 will test the localization of the brain systems underlying the prediction of likely event participants and arguments by comparing the performance of participants with aphasia due to LH damage (either anterior only or extending to posterior areas) to that of unimpaired participants in a visual world experiment. Experiment 2 will use this paradigm to test whether predictions for arguments vs. likely event participants have different properties and are generated by different brain systems. Experiments 3 will monitor the eye movements of both unimpaired adults and adults with aphasia while they are reading to determine which brain systems are quickly sensitive to impossible arguments, in order to test our hypothesis that there are semantically vacuous argument-specific predictions that can be satisfied by any word that results in a possible event interpretation. Experiment 4 will use the same method to investigate whether there is competition between the output of the different brain systems recruited during verb-argument comprehension. Experiment 5 gathers fMRI data from unimpaired participants to provide converging evidence regarding the localization of predictions for arguments and likely event participants and the interpretation of verb- argument combinations. The overarching goal of this work is to improve our understanding of the neurolinguistic processes underlying verb-argument processing and their impairment following brain damage.